Strain-level bacterial typing directly from patient samples using optical DNA mapping

Abstract

Background: Identification of pathogens is crucial to efficiently treat and prevent bacterial infections. However, existing diagnostic techniques are slow or have a too low resolution for well-informed clinical decisions.

Methods: In this study, we have developed an optical DNA mapping-based method for strain-level bacterial typing and simultaneous plasmid characterisation. For the typing, different taxonomical resolutions were examined and cultivated pure Escherichia coli and Klebsiella pneumoniae samples were used for parameter optimization. Finally, the method was applied to mixed bacterial samples and uncultured urine samples from patients with urinary tract infections.

Results: We demonstrate that optical DNA mapping of single DNA molecules can identify Escherichia coli and Klebsiella pneumoniae at the strain level directly from patient samples. At a taxonomic resolution corresponding to E. coli sequence type 131 and K. pneumoniae clonal complex 258 forming distinct groups, the average true positive prediction rates are 94% and 89%, respectively. The single-molecule aspect of the method enables us to identify multiple E. coli strains in polymicrobial samples. Furthermore, by targeting plasmid-borne antibiotic resistance genes with Cas9 restriction, we simultaneously identify the strain or subtype and characterize the corresponding plasmids.

Conclusion: The optical DNA mapping method is accurate and directly applicable to polymicrobial and clinical samples without cultivation. Hence, it has the potential to rapidly provide comprehensive diagnostics information, thereby optimizing early antibiotic treatment and opening up for future precision medicine management.

Fig. 1. Schematic overview of high-resolution optical DNA mapping-based bacterial typing.

a Experimental pipeline. The DNA, extracted via plug-lysis, is labelled with YOYO-1 and netropsin in a single step. The DNA is confined to nanofluidic channels and imaged using a fluorescence microscope, resulting in one experimental intensity profile per analysed DNA fragment. b Data analysis pipeline. Each experimental intensity profile is matched to the reference database of predicted intensity profiles. Next, the database matches are filtered for each experimental profile so that only high-scoring matches remain. Profiles are deemed to match discriminatively if all remaining matches are against a single strain group (SG) at the investigated resolution. Only the results of the discriminative profiles are reported back to the user. The result for the sample is shown as the number of discriminative fragments matching each detected SG and the proportions of all discriminative profiles matching each detected SG.

Fig. 2. Parameter evaluation and typing results for pure cultured E. coli and K. pneumoniae isolates.

a–d The effects of different filtering stringencies (parameter C_diff) on the average true positive rate (TPR) and the average proportion of discriminative profiles were assessed using 25 E. coli samples (a, c) and nine K. pneumoniae samples (b, d). The method was evaluated on the three investigated strain-level taxonomic resolutions – SG_Low, SG_Medium and SG_High – as well as on the species level. e The number of discriminative intensity profiles and true positive rate (TPR) for each cultivated pure E. coli sample for the three investigated taxonomic resolutions, SG_High, SG_Medium and SG_Low. f The number of discriminative intensity profiles and true positive rate (TPR) for each cultivated pure K. pneumoniae sample for the three investigated taxonomic resolutions, SG_High, SG_Medium and SG_Low.

Fig. 3. Typing results for mixes, urine samples, and ST131 samples with ultra-high resolution.

a The distribution of discriminative strain group (SG) matches at SG_Medium-resolution for the two mixed samples, each containing four different E. coli strains in equal proportions. The number of profiles matching discriminatively to each identified SG is indicated next to the chart. b The number of discriminative intensity profiles and true positive rate (TPR) for six non-cultivated urine samples at the three investigated strain-level taxonomic resolutions, SG_High, SG_Medium and SG_Low. c The ST131 branch of the phylogenetic tree (full tree in Supplementary Fig. 1) includes the nine ST131 samples (white) and the closest non-ST131 genome (far left in grey). The five SGs for this ultra-high resolution are marked (yellow-purple). d The number of discriminative intensity profiles and true positive rate (TPR) for the nine ST131 samples plotted at ultra-high resolution (SG_Ultra-High), corresponding to the five SGs of the ST131 branch.

Fig. 4. Simultaneous bacterial typing and plasmid characterization.

a Experimental pipeline. DNA is extracted via plug-lysis and the gene of interest is cut by Cas9 followed by YOYO-1 and netropsin labelling. The labelled DNA sample, containing both chromosomal and plasmid DNA, is confined to nanofluidic channels and imaged using a fluorescence microscope, resulting in one experimental intensity profile per DNA fragment. b Bacterial typing using profiles longer than 250 pixels (~135 kb). The bar charts show the true positive rate (TPR) and the number of discriminative intensity profiles for samples P1, P2 and P3 at the three strain group (SG) resolutions, SG_High, SG_Medium and SG_Low. c The intensity profiles for the consensuses from the plasmids from P1, P2, and P3 shifted vertically for clarity. d Detection of plasmid-borne bla_CTX-M group 1 and 9 genes. In the circular plot, each ribbon represents individual DNA fragments and their brightness corresponds to the intensity profile. The profiles have been aligned and the consensus of all the profiles is included as the outermost ribbon. Linearization at the same position indicates a cut by Cas9 and verifies the presence of the target gene (see examples of randomly linearized in Supplementary Fig. 3). The three circular plots show linearization by Cas9 for the 99 kb plasmid in sample P1 (12/12 profiles linearized at the same position), 121 kb plasmid in sample P2 (12/13) and for the 66 kb plasmid in sample P3 (26/28).